Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[SPARK-33287][SS][UI]Expose state custom metrics information on SS UI #30336

Closed
wants to merge 15 commits into from
Closed

[SPARK-33287][SS][UI]Expose state custom metrics information on SS UI #30336

Show file tree
Hide file tree
Changes from 13 commits
Commits
Show all changes
15 commits
Select commit Hold shift + click to select a range
2eefeb2
[SPARK-33287][SS][UI]Expose state custom metrics information on SS UI
gaborgsomogyi Nov 10, 2020
6654934
Partial review fixes
gaborgsomogyi Nov 13, 2020
f11bc22
Switch to enable flag
gaborgsomogyi Nov 13, 2020
19e2da1
Cleanup
gaborgsomogyi Nov 13, 2020
34cb99f
Move config to StaticSQLConf + nit fixes
gaborgsomogyi Nov 16, 2020
a788234
Merge branch 'master' into SPARK-33287
gaborgsomogyi Nov 16, 2020
7787764
Indent fix
gaborgsomogyi Nov 17, 2020
791011c
Add unit and description
gaborgsomogyi Nov 18, 2020
5700455
Test fix
gaborgsomogyi Nov 18, 2020
d43b1b9
Unit default value
gaborgsomogyi Nov 19, 2020
bc83741
Merge branch 'master' into SPARK-33287
gaborgsomogyi Nov 20, 2020
9a87255
Remove unit + rename a variable
gaborgsomogyi Nov 20, 2020
864062f
Merge branch 'master' into SPARK-33287
gaborgsomogyi Nov 21, 2020
115de28
Simplify enabled custom metrics check
gaborgsomogyi Nov 24, 2020
e2f1594
Ignore case compare
gaborgsomogyi Nov 24, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 12 additions & 0 deletions sql/catalyst/src/main/scala/org/apache/spark/sql/internal/StaticSQLConf.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -249,4 +249,16 @@ object StaticSQLConf {
.version("3.1.0")
.timeConf(TimeUnit.SECONDS)
.createWithDefault(-1)

val ENABLED_STREAMING_UI_CUSTOM_METRIC_LIST =
buildStaticConf("spark.sql.streaming.ui.enabledCustomMetricList")
.internal()
.doc("Configures a list of custom metrics on Structured Streaming UI, which are enabled. " +
"The list contains the name of the custom metrics separated by comma. In aggregation" +
"only sum used. The list of supported custom metrics is state store provider specific " +
"and it can be found out for example from query progress log entry.")
.version("3.1.0")
.stringConf
.toSequence
.createWithDefault(Nil)
}
137 changes: 99 additions & 38 deletions sql/core/src/main/scala/org/apache/spark/sql/streaming/ui/StreamingQueryStatisticsPage.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,15 +22,25 @@ import java.lang.{Long => JLong}
import java.util.UUID
import javax.servlet.http.HttpServletRequest

import scala.collection.JavaConverters._
import scala.xml.{Node, NodeBuffer, Unparsed}

import org.apache.spark.internal.Logging
import org.apache.spark.sql.execution.streaming.state.StateStoreProvider
import org.apache.spark.sql.internal.SQLConf.STATE_STORE_PROVIDER_CLASS
import org.apache.spark.sql.internal.StaticSQLConf.ENABLED_STREAMING_UI_CUSTOM_METRIC_LIST
import org.apache.spark.sql.streaming.ui.UIUtils._
import org.apache.spark.ui.{GraphUIData, JsCollector, UIUtils => SparkUIUtils, WebUIPage}

private[ui] class StreamingQueryStatisticsPage(parent: StreamingQueryTab)
extends WebUIPage("statistics") with Logging {

// State store provider implementation mustn't do any heavyweight initialiation in constructor
// but in its init method.
private val supportedCustomMetrics = StateStoreProvider.create(
parent.parent.conf.get(STATE_STORE_PROVIDER_CLASS)).supportedCustomMetrics
logDebug(s"Supported custom metrics: $supportedCustomMetrics")
gaborgsomogyi marked this conversation as resolved.
Show resolved Hide resolved

def generateLoadResources(request: HttpServletRequest): Seq[Node] = {
// scalastyle:off
<script src={SparkUIUtils.prependBaseUri(request, "/static/d3.min.js")}></script>
Expand Down Expand Up @@ -199,49 +209,100 @@ private[ui] class StreamingQueryStatisticsPage(parent: StreamingQueryTab)
"records")
graphUIDataForNumRowsDroppedByWatermark.generateDataJs(jsCollector)

// scalastyle:off
<tr>
<td style="vertical-align: middle;">
<div style="width: 160px;">
<div><strong>Aggregated Number Of Total State Rows {SparkUIUtils.tooltip("Aggregated number of total state rows.", "right")}</strong></div>
</div>
</td>
<td class={"aggregated-num-total-state-rows-timeline"}>{graphUIDataForNumberTotalRows.generateTimelineHtml(jsCollector)}</td>
<td class={"aggregated-num-total-state-rows-histogram"}>{graphUIDataForNumberTotalRows.generateHistogramHtml(jsCollector)}</td>
</tr>
<tr>
<td style="vertical-align: middle;">
<div style="width: 160px;">
<div><strong>Aggregated Number Of Updated State Rows {SparkUIUtils.tooltip("Aggregated number of updated state rows.", "right")}</strong></div>
</div>
</td>
<td class={"aggregated-num-updated-state-rows-timeline"}>{graphUIDataForNumberUpdatedRows.generateTimelineHtml(jsCollector)}</td>
<td class={"aggregated-num-updated-state-rows-histogram"}>{graphUIDataForNumberUpdatedRows.generateHistogramHtml(jsCollector)}</td>
</tr>
<tr>
<td style="vertical-align: middle;">
<div style="width: 160px;">
<div><strong>Aggregated State Memory Used In Bytes {SparkUIUtils.tooltip("Aggregated state memory used in bytes.", "right")}</strong></div>
</div>
</td>
<td class={"aggregated-state-memory-used-bytes-timeline"}>{graphUIDataForMemoryUsedBytes.generateTimelineHtml(jsCollector)}</td>
<td class={"aggregated-state-memory-used-bytes-histogram"}>{graphUIDataForMemoryUsedBytes.generateHistogramHtml(jsCollector)}</td>
</tr>
<tr>
<td style="vertical-align: middle;">
<div style="width: 160px;">
<div><strong>Aggregated Number Of Rows Dropped By Watermark {SparkUIUtils.tooltip("Accumulates all input rows being dropped in stateful operators by watermark. 'Inputs' are relative to operators.", "right")}</strong></div>
</div>
</td>
<td class={"aggregated-num-rows-dropped-by-watermark-timeline"}>{graphUIDataForNumRowsDroppedByWatermark.generateTimelineHtml(jsCollector)}</td>
<td class={"aggregated-num-rows-dropped-by-watermark-histogram"}>{graphUIDataForNumRowsDroppedByWatermark.generateHistogramHtml(jsCollector)}</td>
</tr>
// scalastyle:on
val result =
// scalastyle:off
<tr>
<td style="vertical-align: middle;">
<div style="width: 160px;">
<div><strong>Aggregated Number Of Total State Rows {SparkUIUtils.tooltip("Aggregated number of total state rows.", "right")}</strong></div>
</div>
</td>
<td class={"aggregated-num-total-state-rows-timeline"}>{graphUIDataForNumberTotalRows.generateTimelineHtml(jsCollector)}</td>
<td class={"aggregated-num-total-state-rows-histogram"}>{graphUIDataForNumberTotalRows.generateHistogramHtml(jsCollector)}</td>
</tr>
<tr>
<td style="vertical-align: middle;">
<div style="width: 160px;">
<div><strong>Aggregated Number Of Updated State Rows {SparkUIUtils.tooltip("Aggregated number of updated state rows.", "right")}</strong></div>
</div>
</td>
<td class={"aggregated-num-updated-state-rows-timeline"}>{graphUIDataForNumberUpdatedRows.generateTimelineHtml(jsCollector)}</td>
<td class={"aggregated-num-updated-state-rows-histogram"}>{graphUIDataForNumberUpdatedRows.generateHistogramHtml(jsCollector)}</td>
</tr>
<tr>
<td style="vertical-align: middle;">
<div style="width: 160px;">
<div><strong>Aggregated State Memory Used In Bytes {SparkUIUtils.tooltip("Aggregated state memory used in bytes.", "right")}</strong></div>
</div>
</td>
<td class={"aggregated-state-memory-used-bytes-timeline"}>{graphUIDataForMemoryUsedBytes.generateTimelineHtml(jsCollector)}</td>
<td class={"aggregated-state-memory-used-bytes-histogram"}>{graphUIDataForMemoryUsedBytes.generateHistogramHtml(jsCollector)}</td>
</tr>
<tr>
<td style="vertical-align: middle;">
<div style="width: 160px;">
<div><strong>Aggregated Number Of Rows Dropped By Watermark {SparkUIUtils.tooltip("Accumulates all input rows being dropped in stateful operators by watermark. 'Inputs' are relative to operators.", "right")}</strong></div>
</div>
</td>
<td class={"aggregated-num-rows-dropped-by-watermark-timeline"}>{graphUIDataForNumRowsDroppedByWatermark.generateTimelineHtml(jsCollector)}</td>
<td class={"aggregated-num-rows-dropped-by-watermark-histogram"}>{graphUIDataForNumRowsDroppedByWatermark.generateHistogramHtml(jsCollector)}</td>
</tr>
// scalastyle:on

result ++= generateAggregatedCustomMetrics(query, minBatchTime, maxBatchTime, jsCollector)
result
} else {
new NodeBuffer()
}
}

def generateAggregatedCustomMetrics(
query: StreamingQueryUIData,
minBatchTime: Long,
maxBatchTime: Long,
jsCollector: JsCollector): NodeBuffer = {
val result: NodeBuffer = new NodeBuffer

// This is made sure on caller side but put it here to be defensive
require(query.lastProgress.stateOperators.nonEmpty)
val enabledCustomMetrics = parent.parent.conf.get(ENABLED_STREAMING_UI_CUSTOM_METRIC_LIST)
gaborgsomogyi marked this conversation as resolved.
Show resolved Hide resolved
logDebug(s"Enabled custom metrics: $enabledCustomMetrics")
query.lastProgress.stateOperators.head.customMetrics.keySet().asScala
gaborgsomogyi marked this conversation as resolved.
Show resolved Hide resolved
.filter(enabledCustomMetrics.contains(_)).map { metricName =>
gaborgsomogyi marked this conversation as resolved.
Show resolved Hide resolved
val data = query.recentProgress.map(p => (parseProgressTimestamp(p.timestamp),
p.stateOperators.map(_.customMetrics.get(metricName).toDouble).sum))
val max = data.maxBy(_._2)._2
val metric = supportedCustomMetrics.find(_.name == metricName).get

val graphUIData =
new GraphUIData(
s"aggregated-$metricName-timeline",
s"aggregated-$metricName-histogram",
data,
minBatchTime,
maxBatchTime,
0,
max,
"")
gaborgsomogyi marked this conversation as resolved.
Show resolved Hide resolved
graphUIData.generateDataJs(jsCollector)

result ++=
// scalastyle:off
<tr>
<td style="vertical-align: middle;">
<div style="width: 240px;">
<div><strong>Aggregated Custom Metric {s"$metricName"} {SparkUIUtils.tooltip(metric.desc, "right")}</strong></div>
</div>
</td>
<td class={s"aggregated-$metricName-timeline"}>{graphUIData.generateTimelineHtml(jsCollector)}</td>
<td class={s"aggregated-$metricName-histogram"}>{graphUIData.generateHistogramHtml(jsCollector)}</td>
</tr>
// scalastyle:on
}

result
}

def generateStatTable(query: StreamingQueryUIData): Seq[Node] = {
val batchToTimestamps = withNoProgress(query,
query.recentProgress.map(p => (p.batchId, parseProgressTimestamp(p.timestamp))),
Expand Down
5 changes: 5 additions & 0 deletions sql/core/src/test/scala/org/apache/spark/sql/streaming/ui/StreamingQueryPageSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,8 +24,10 @@ import org.mockito.Mockito.{mock, when, RETURNS_SMART_NULLS}
import org.scalatest.BeforeAndAfter
import scala.xml.Node

import org.apache.spark.SparkConf
import org.apache.spark.sql.streaming.StreamingQueryProgress
import org.apache.spark.sql.test.SharedSparkSession
import org.apache.spark.ui.SparkUI

class StreamingQueryPageSuite extends SharedSparkSession with BeforeAndAfter {

Expand Down Expand Up @@ -65,10 +67,13 @@ class StreamingQueryPageSuite extends SharedSparkSession with BeforeAndAfter {
val request = mock(classOf[HttpServletRequest])
val tab = mock(classOf[StreamingQueryTab], RETURNS_SMART_NULLS)
val statusListener = mock(classOf[StreamingQueryStatusListener], RETURNS_SMART_NULLS)
val ui = mock(classOf[SparkUI])
when(request.getParameter("id")).thenReturn(id.toString)
when(tab.appName).thenReturn("testing")
when(tab.headerTabs).thenReturn(Seq.empty)
when(tab.statusListener).thenReturn(statusListener)
when(ui.conf).thenReturn(new SparkConf())
when(tab.parent).thenReturn(ui)

val streamQuery = createStreamQueryUIData(id)
when(statusListener.allQueryStatus).thenReturn(Seq(streamQuery))
Expand Down
6 changes: 6 additions & 0 deletions sql/core/src/test/scala/org/apache/spark/sql/streaming/ui/UISeleniumSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,6 +31,7 @@ import org.apache.spark.internal.config.UI.{UI_ENABLED, UI_PORT}
import org.apache.spark.sql.LocalSparkSession.withSparkSession
import org.apache.spark.sql.SparkSession
import org.apache.spark.sql.catalyst.util.quietly
import org.apache.spark.sql.internal.StaticSQLConf.ENABLED_STREAMING_UI_CUSTOM_METRIC_LIST
import org.apache.spark.sql.streaming.StreamingQueryException
import org.apache.spark.ui.SparkUICssErrorHandler

Expand All @@ -53,6 +54,7 @@ class UISeleniumSuite extends SparkFunSuite with WebBrowser with Matchers with B
.setAppName("ui-test")
.set(UI_ENABLED, true)
.set(UI_PORT, 0)
.set(ENABLED_STREAMING_UI_CUSTOM_METRIC_LIST, Seq("stateOnCurrentVersionSizeBytes"))
additionalConfs.foreach { case (k, v) => conf.set(k, v) }
val spark = SparkSession.builder().master(master).config(conf).getOrCreate()
assert(spark.sparkContext.ui.isDefined)
Expand Down Expand Up @@ -140,6 +142,10 @@ class UISeleniumSuite extends SparkFunSuite with WebBrowser with Matchers with B
summaryText should contain ("Aggregated Number Of Updated State Rows (?)")
summaryText should contain ("Aggregated State Memory Used In Bytes (?)")
summaryText should contain ("Aggregated Number Of Rows Dropped By Watermark (?)")
summaryText should contain ("Aggregated Custom Metric stateOnCurrentVersionSizeBytes" +
" (?)")
summaryText should not contain ("Aggregated Custom Metric loadedMapCacheHitCount (?)")
summaryText should not contain ("Aggregated Custom Metric loadedMapCacheMissCount (?)")
}
} finally {
spark.streams.active.foreach(_.stop())
Expand Down